linux/Documentation/filesystems/proc.txt
<<
>>
Prefs
   1------------------------------------------------------------------------------
   2                       T H E  /proc   F I L E S Y S T E M
   3------------------------------------------------------------------------------
   4/proc/sys         Terrehon Bowden <terrehon@pacbell.net>        October 7 1999
   5                  Bodo Bauer <bb@ricochet.net>
   6
   72.4.x update      Jorge Nerin <comandante@zaralinux.com>      November 14 2000
   8move /proc/sys    Shen Feng <shen@cn.fujitsu.com>                 April 1 2009
   9------------------------------------------------------------------------------
  10Version 1.3                                              Kernel version 2.2.12
  11                                              Kernel version 2.4.0-test11-pre4
  12------------------------------------------------------------------------------
  13fixes/update part 1.1  Stefani Seibold <stefani@seibold.net>       June 9 2009
  14
  15Table of Contents
  16-----------------
  17
  18  0     Preface
  19  0.1   Introduction/Credits
  20  0.2   Legal Stuff
  21
  22  1     Collecting System Information
  23  1.1   Process-Specific Subdirectories
  24  1.2   Kernel data
  25  1.3   IDE devices in /proc/ide
  26  1.4   Networking info in /proc/net
  27  1.5   SCSI info
  28  1.6   Parallel port info in /proc/parport
  29  1.7   TTY info in /proc/tty
  30  1.8   Miscellaneous kernel statistics in /proc/stat
  31  1.9   Ext4 file system parameters
  32
  33  2     Modifying System Parameters
  34
  35  3     Per-Process Parameters
  36  3.1   /proc/<pid>/oom_adj & /proc/<pid>/oom_score_adj - Adjust the oom-killer
  37                                                                score
  38  3.2   /proc/<pid>/oom_score - Display current oom-killer score
  39  3.3   /proc/<pid>/io - Display the IO accounting fields
  40  3.4   /proc/<pid>/coredump_filter - Core dump filtering settings
  41  3.5   /proc/<pid>/mountinfo - Information about mounts
  42  3.6   /proc/<pid>/comm  & /proc/<pid>/task/<tid>/comm
  43  3.7   /proc/<pid>/task/<tid>/children - Information about task children
  44  3.8   /proc/<pid>/fdinfo/<fd> - Information about opened file
  45  3.9   /proc/<pid>/map_files - Information about memory mapped files
  46  3.10  /proc/<pid>/timerslack_ns - Task timerslack value
  47
  48  4     Configuring procfs
  49  4.1   Mount options
  50
  51------------------------------------------------------------------------------
  52Preface
  53------------------------------------------------------------------------------
  54
  550.1 Introduction/Credits
  56------------------------
  57
  58This documentation is  part of a soon (or  so we hope) to be  released book on
  59the SuSE  Linux distribution. As  there is  no complete documentation  for the
  60/proc file system and we've used  many freely available sources to write these
  61chapters, it  seems only fair  to give the work  back to the  Linux community.
  62This work is  based on the 2.2.*  kernel version and the  upcoming 2.4.*. I'm
  63afraid it's still far from complete, but we  hope it will be useful. As far as
  64we know, it is the first 'all-in-one' document about the /proc file system. It
  65is focused  on the Intel  x86 hardware,  so if you  are looking for  PPC, ARM,
  66SPARC, AXP, etc., features, you probably  won't find what you are looking for.
  67It also only covers IPv4 networking, not IPv6 nor other protocols - sorry. But
  68additions and patches  are welcome and will  be added to this  document if you
  69mail them to Bodo.
  70
  71We'd like  to  thank Alan Cox, Rik van Riel, and Alexey Kuznetsov and a lot of
  72other people for help compiling this documentation. We'd also like to extend a
  73special thank  you to Andi Kleen for documentation, which we relied on heavily
  74to create  this  document,  as well as the additional information he provided.
  75Thanks to  everybody  else  who contributed source or docs to the Linux kernel
  76and helped create a great piece of software... :)
  77
  78If you  have  any comments, corrections or additions, please don't hesitate to
  79contact Bodo  Bauer  at  bb@ricochet.net.  We'll  be happy to add them to this
  80document.
  81
  82The   latest   version    of   this   document   is    available   online   at
  83http://tldp.org/LDP/Linux-Filesystem-Hierarchy/html/proc.html
  84
  85If  the above  direction does  not works  for you,  you could  try the  kernel
  86mailing  list  at  linux-kernel@vger.kernel.org  and/or try  to  reach  me  at
  87comandante@zaralinux.com.
  88
  890.2 Legal Stuff
  90---------------
  91
  92We don't  guarantee  the  correctness  of this document, and if you come to us
  93complaining about  how  you  screwed  up  your  system  because  of  incorrect
  94documentation, we won't feel responsible...
  95
  96------------------------------------------------------------------------------
  97CHAPTER 1: COLLECTING SYSTEM INFORMATION
  98------------------------------------------------------------------------------
  99
 100------------------------------------------------------------------------------
 101In This Chapter
 102------------------------------------------------------------------------------
 103* Investigating  the  properties  of  the  pseudo  file  system  /proc and its
 104  ability to provide information on the running Linux system
 105* Examining /proc's structure
 106* Uncovering  various  information  about the kernel and the processes running
 107  on the system
 108------------------------------------------------------------------------------
 109
 110
 111The proc  file  system acts as an interface to internal data structures in the
 112kernel. It  can  be  used to obtain information about the system and to change
 113certain kernel parameters at runtime (sysctl).
 114
 115First, we'll  take  a  look  at the read-only parts of /proc. In Chapter 2, we
 116show you how you can use /proc/sys to change settings.
 117
 1181.1 Process-Specific Subdirectories
 119-----------------------------------
 120
 121The directory  /proc  contains  (among other things) one subdirectory for each
 122process running on the system, which is named after the process ID (PID).
 123
 124The link  self  points  to  the  process reading the file system. Each process
 125subdirectory has the entries listed in Table 1-1.
 126
 127
 128Table 1-1: Process specific entries in /proc
 129..............................................................................
 130 File           Content
 131 clear_refs     Clears page referenced bits shown in smaps output
 132 cmdline        Command line arguments
 133 cpu            Current and last cpu in which it was executed   (2.4)(smp)
 134 cwd            Link to the current working directory
 135 environ        Values of environment variables
 136 exe            Link to the executable of this process
 137 fd             Directory, which contains all file descriptors
 138 maps           Memory maps to executables and library files    (2.4)
 139 mem            Memory held by this process
 140 root           Link to the root directory of this process
 141 stat           Process status
 142 statm          Process memory status information
 143 status         Process status in human readable form
 144 wchan          Present with CONFIG_KALLSYMS=y: it shows the kernel function
 145                symbol the task is blocked in - or "0" if not blocked.
 146 pagemap        Page table
 147 stack          Report full stack trace, enable via CONFIG_STACKTRACE
 148 smaps          a extension based on maps, showing the memory consumption of
 149                each mapping and flags associated with it
 150 numa_maps      an extension based on maps, showing the memory locality and
 151                binding policy as well as mem usage (in pages) of each mapping.
 152..............................................................................
 153
 154For example, to get the status information of a process, all you have to do is
 155read the file /proc/PID/status:
 156
 157  >cat /proc/self/status
 158  Name:   cat
 159  State:  R (running)
 160  Tgid:   5452
 161  Pid:    5452
 162  PPid:   743
 163  TracerPid:      0                                             (2.4)
 164  Uid:    501     501     501     501
 165  Gid:    100     100     100     100
 166  FDSize: 256
 167  Groups: 100 14 16
 168  VmPeak:     5004 kB
 169  VmSize:     5004 kB
 170  VmLck:         0 kB
 171  VmHWM:       476 kB
 172  VmRSS:       476 kB
 173  RssAnon:             352 kB
 174  RssFile:             120 kB
 175  RssShmem:              4 kB
 176  VmData:      156 kB
 177  VmStk:        88 kB
 178  VmExe:        68 kB
 179  VmLib:      1412 kB
 180  VmPTE:        20 kb
 181  VmSwap:        0 kB
 182  HugetlbPages:          0 kB
 183  Threads:        1
 184  SigQ:   0/28578
 185  SigPnd: 0000000000000000
 186  ShdPnd: 0000000000000000
 187  SigBlk: 0000000000000000
 188  SigIgn: 0000000000000000
 189  SigCgt: 0000000000000000
 190  CapInh: 00000000fffffeff
 191  CapPrm: 0000000000000000
 192  CapEff: 0000000000000000
 193  CapBnd: ffffffffffffffff
 194  Seccomp:        0
 195  voluntary_ctxt_switches:        0
 196  nonvoluntary_ctxt_switches:     1
 197
 198This shows you nearly the same information you would get if you viewed it with
 199the ps  command.  In  fact,  ps  uses  the  proc  file  system  to  obtain its
 200information.  But you get a more detailed  view of the  process by reading the
 201file /proc/PID/status. It fields are described in table 1-2.
 202
 203The  statm  file  contains  more  detailed  information about the process
 204memory usage. Its seven fields are explained in Table 1-3.  The stat file
 205contains details information about the process itself.  Its fields are
 206explained in Table 1-4.
 207
 208(for SMP CONFIG users)
 209For making accounting scalable, RSS related information are handled in an
 210asynchronous manner and the value may not be very precise. To see a precise
 211snapshot of a moment, you can see /proc/<pid>/smaps file and scan page table.
 212It's slow but very precise.
 213
 214Table 1-2: Contents of the status files (as of 4.1)
 215..............................................................................
 216 Field                       Content
 217 Name                        filename of the executable
 218 State                       state (R is running, S is sleeping, D is sleeping
 219                             in an uninterruptible wait, Z is zombie,
 220                             T is traced or stopped)
 221 Tgid                        thread group ID
 222 Ngid                        NUMA group ID (0 if none)
 223 Pid                         process id
 224 PPid                        process id of the parent process
 225 TracerPid                   PID of process tracing this process (0 if not)
 226 Uid                         Real, effective, saved set, and  file system UIDs
 227 Gid                         Real, effective, saved set, and  file system GIDs
 228 FDSize                      number of file descriptor slots currently allocated
 229 Groups                      supplementary group list
 230 NStgid                      descendant namespace thread group ID hierarchy
 231 NSpid                       descendant namespace process ID hierarchy
 232 NSpgid                      descendant namespace process group ID hierarchy
 233 NSsid                       descendant namespace session ID hierarchy
 234 VmPeak                      peak virtual memory size
 235 VmSize                      total program size
 236 VmLck                       locked memory size
 237 VmHWM                       peak resident set size ("high water mark")
 238 VmRSS                       size of memory portions. It contains the three
 239                             following parts (VmRSS = RssAnon + RssFile + RssShmem)
 240 RssAnon                     size of resident anonymous memory
 241 RssFile                     size of resident file mappings
 242 RssShmem                    size of resident shmem memory (includes SysV shm,
 243                             mapping of tmpfs and shared anonymous mappings)
 244 VmData                      size of private data segments
 245 VmStk                       size of stack segments
 246 VmExe                       size of text segment
 247 VmLib                       size of shared library code
 248 VmPTE                       size of page table entries
 249 VmPMD                       size of second level page tables
 250 VmSwap                      amount of swap used by anonymous private data
 251                             (shmem swap usage is not included)
 252 HugetlbPages                size of hugetlb memory portions
 253 Threads                     number of threads
 254 SigQ                        number of signals queued/max. number for queue
 255 SigPnd                      bitmap of pending signals for the thread
 256 ShdPnd                      bitmap of shared pending signals for the process
 257 SigBlk                      bitmap of blocked signals
 258 SigIgn                      bitmap of ignored signals
 259 SigCgt                      bitmap of caught signals
 260 CapInh                      bitmap of inheritable capabilities
 261 CapPrm                      bitmap of permitted capabilities
 262 CapEff                      bitmap of effective capabilities
 263 CapBnd                      bitmap of capabilities bounding set
 264 Seccomp                     seccomp mode, like prctl(PR_GET_SECCOMP, ...)
 265 Cpus_allowed                mask of CPUs on which this process may run
 266 Cpus_allowed_list           Same as previous, but in "list format"
 267 Mems_allowed                mask of memory nodes allowed to this process
 268 Mems_allowed_list           Same as previous, but in "list format"
 269 voluntary_ctxt_switches     number of voluntary context switches
 270 nonvoluntary_ctxt_switches  number of non voluntary context switches
 271..............................................................................
 272
 273Table 1-3: Contents of the statm files (as of 2.6.8-rc3)
 274..............................................................................
 275 Field    Content
 276 size     total program size (pages)            (same as VmSize in status)
 277 resident size of memory portions (pages)       (same as VmRSS in status)
 278 shared   number of pages that are shared       (i.e. backed by a file, same
 279                                                as RssFile+RssShmem in status)
 280 trs      number of pages that are 'code'       (not including libs; broken,
 281                                                        includes data segment)
 282 lrs      number of pages of library            (always 0 on 2.6)
 283 drs      number of pages of data/stack         (including libs; broken,
 284                                                        includes library text)
 285 dt       number of dirty pages                 (always 0 on 2.6)
 286..............................................................................
 287
 288
 289Table 1-4: Contents of the stat files (as of 2.6.30-rc7)
 290..............................................................................
 291 Field          Content
 292  pid           process id
 293  tcomm         filename of the executable
 294  state         state (R is running, S is sleeping, D is sleeping in an
 295                uninterruptible wait, Z is zombie, T is traced or stopped)
 296  ppid          process id of the parent process
 297  pgrp          pgrp of the process
 298  sid           session id
 299  tty_nr        tty the process uses
 300  tty_pgrp      pgrp of the tty
 301  flags         task flags
 302  min_flt       number of minor faults
 303  cmin_flt      number of minor faults with child's
 304  maj_flt       number of major faults
 305  cmaj_flt      number of major faults with child's
 306  utime         user mode jiffies
 307  stime         kernel mode jiffies
 308  cutime        user mode jiffies with child's
 309  cstime        kernel mode jiffies with child's
 310  priority      priority level
 311  nice          nice level
 312  num_threads   number of threads
 313  it_real_value (obsolete, always 0)
 314  start_time    time the process started after system boot
 315  vsize         virtual memory size
 316  rss           resident set memory size
 317  rsslim        current limit in bytes on the rss
 318  start_code    address above which program text can run
 319  end_code      address below which program text can run
 320  start_stack   address of the start of the main process stack
 321  esp           current value of ESP
 322  eip           current value of EIP
 323  pending       bitmap of pending signals
 324  blocked       bitmap of blocked signals
 325  sigign        bitmap of ignored signals
 326  sigcatch      bitmap of caught signals
 327  0             (place holder, used to be the wchan address, use /proc/PID/wchan instead)
 328  0             (place holder)
 329  0             (place holder)
 330  exit_signal   signal to send to parent thread on exit
 331  task_cpu      which CPU the task is scheduled on
 332  rt_priority   realtime priority
 333  policy        scheduling policy (man sched_setscheduler)
 334  blkio_ticks   time spent waiting for block IO
 335  gtime         guest time of the task in jiffies
 336  cgtime        guest time of the task children in jiffies
 337  start_data    address above which program data+bss is placed
 338  end_data      address below which program data+bss is placed
 339  start_brk     address above which program heap can be expanded with brk()
 340  arg_start     address above which program command line is placed
 341  arg_end       address below which program command line is placed
 342  env_start     address above which program environment is placed
 343  env_end       address below which program environment is placed
 344  exit_code     the thread's exit_code in the form reported by the waitpid system call
 345..............................................................................
 346
 347The /proc/PID/maps file containing the currently mapped memory regions and
 348their access permissions.
 349
 350The format is:
 351
 352address           perms offset  dev   inode      pathname
 353
 35408048000-08049000 r-xp 00000000 03:00 8312       /opt/test
 35508049000-0804a000 rw-p 00001000 03:00 8312       /opt/test
 3560804a000-0806b000 rw-p 00000000 00:00 0          [heap]
 357a7cb1000-a7cb2000 ---p 00000000 00:00 0
 358a7cb2000-a7eb2000 rw-p 00000000 00:00 0
 359a7eb2000-a7eb3000 ---p 00000000 00:00 0
 360a7eb3000-a7ed5000 rw-p 00000000 00:00 0
 361a7ed5000-a8008000 r-xp 00000000 03:00 4222       /lib/libc.so.6
 362a8008000-a800a000 r--p 00133000 03:00 4222       /lib/libc.so.6
 363a800a000-a800b000 rw-p 00135000 03:00 4222       /lib/libc.so.6
 364a800b000-a800e000 rw-p 00000000 00:00 0
 365a800e000-a8022000 r-xp 00000000 03:00 14462      /lib/libpthread.so.0
 366a8022000-a8023000 r--p 00013000 03:00 14462      /lib/libpthread.so.0
 367a8023000-a8024000 rw-p 00014000 03:00 14462      /lib/libpthread.so.0
 368a8024000-a8027000 rw-p 00000000 00:00 0
 369a8027000-a8043000 r-xp 00000000 03:00 8317       /lib/ld-linux.so.2
 370a8043000-a8044000 r--p 0001b000 03:00 8317       /lib/ld-linux.so.2
 371a8044000-a8045000 rw-p 0001c000 03:00 8317       /lib/ld-linux.so.2
 372aff35000-aff4a000 rw-p 00000000 00:00 0          [stack]
 373ffffe000-fffff000 r-xp 00000000 00:00 0          [vdso]
 374
 375where "address" is the address space in the process that it occupies, "perms"
 376is a set of permissions:
 377
 378 r = read
 379 w = write
 380 x = execute
 381 s = shared
 382 p = private (copy on write)
 383
 384"offset" is the offset into the mapping, "dev" is the device (major:minor), and
 385"inode" is the inode  on that device.  0 indicates that  no inode is associated
 386with the memory region, as the case would be with BSS (uninitialized data).
 387The "pathname" shows the name associated file for this mapping.  If the mapping
 388is not associated with a file:
 389
 390 [heap]                   = the heap of the program
 391 [stack]                  = the stack of the main process
 392 [vdso]                   = the "virtual dynamic shared object",
 393                            the kernel system call handler
 394
 395 or if empty, the mapping is anonymous.
 396
 397The /proc/PID/task/TID/maps is a view of the virtual memory from the viewpoint
 398of the individual tasks of a process. In this file you will see a mapping marked
 399as [stack] if that task sees it as a stack. Hence, for the example above, the
 400task-level map, i.e. /proc/PID/task/TID/maps for thread 1001 will look like this:
 401
 40208048000-08049000 r-xp 00000000 03:00 8312       /opt/test
 40308049000-0804a000 rw-p 00001000 03:00 8312       /opt/test
 4040804a000-0806b000 rw-p 00000000 00:00 0          [heap]
 405a7cb1000-a7cb2000 ---p 00000000 00:00 0
 406a7cb2000-a7eb2000 rw-p 00000000 00:00 0
 407a7eb2000-a7eb3000 ---p 00000000 00:00 0
 408a7eb3000-a7ed5000 rw-p 00000000 00:00 0          [stack]
 409a7ed5000-a8008000 r-xp 00000000 03:00 4222       /lib/libc.so.6
 410a8008000-a800a000 r--p 00133000 03:00 4222       /lib/libc.so.6
 411a800a000-a800b000 rw-p 00135000 03:00 4222       /lib/libc.so.6
 412a800b000-a800e000 rw-p 00000000 00:00 0
 413a800e000-a8022000 r-xp 00000000 03:00 14462      /lib/libpthread.so.0
 414a8022000-a8023000 r--p 00013000 03:00 14462      /lib/libpthread.so.0
 415a8023000-a8024000 rw-p 00014000 03:00 14462      /lib/libpthread.so.0
 416a8024000-a8027000 rw-p 00000000 00:00 0
 417a8027000-a8043000 r-xp 00000000 03:00 8317       /lib/ld-linux.so.2
 418a8043000-a8044000 r--p 0001b000 03:00 8317       /lib/ld-linux.so.2
 419a8044000-a8045000 rw-p 0001c000 03:00 8317       /lib/ld-linux.so.2
 420aff35000-aff4a000 rw-p 00000000 00:00 0
 421ffffe000-fffff000 r-xp 00000000 00:00 0          [vdso]
 422
 423The /proc/PID/smaps is an extension based on maps, showing the memory
 424consumption for each of the process's mappings. For each of mappings there
 425is a series of lines such as the following:
 426
 42708048000-080bc000 r-xp 00000000 03:02 13130      /bin/bash
 428Size:               1084 kB
 429Rss:                 892 kB
 430Pss:                 374 kB
 431Shared_Clean:        892 kB
 432Shared_Dirty:          0 kB
 433Private_Clean:         0 kB
 434Private_Dirty:         0 kB
 435Referenced:          892 kB
 436Anonymous:             0 kB
 437AnonHugePages:         0 kB
 438Shared_Hugetlb:        0 kB
 439Private_Hugetlb:       0 kB
 440Swap:                  0 kB
 441SwapPss:               0 kB
 442KernelPageSize:        4 kB
 443MMUPageSize:           4 kB
 444Locked:                0 kB
 445VmFlags: rd ex mr mw me dw
 446
 447the first of these lines shows the same information as is displayed for the
 448mapping in /proc/PID/maps.  The remaining lines show the size of the mapping
 449(size), the amount of the mapping that is currently resident in RAM (RSS), the
 450process' proportional share of this mapping (PSS), the number of clean and
 451dirty private pages in the mapping.
 452
 453The "proportional set size" (PSS) of a process is the count of pages it has
 454in memory, where each page is divided by the number of processes sharing it.
 455So if a process has 1000 pages all to itself, and 1000 shared with one other
 456process, its PSS will be 1500.
 457Note that even a page which is part of a MAP_SHARED mapping, but has only
 458a single pte mapped, i.e.  is currently used by only one process, is accounted
 459as private and not as shared.
 460"Referenced" indicates the amount of memory currently marked as referenced or
 461accessed.
 462"Anonymous" shows the amount of memory that does not belong to any file.  Even
 463a mapping associated with a file may contain anonymous pages: when MAP_PRIVATE
 464and a page is modified, the file page is replaced by a private anonymous copy.
 465"AnonHugePages" shows the ammount of memory backed by transparent hugepage.
 466"Shared_Hugetlb" and "Private_Hugetlb" show the ammounts of memory backed by
 467hugetlbfs page which is *not* counted in "RSS" or "PSS" field for historical
 468reasons. And these are not included in {Shared,Private}_{Clean,Dirty} field.
 469"Swap" shows how much would-be-anonymous memory is also used, but out on swap.
 470For shmem mappings, "Swap" includes also the size of the mapped (and not
 471replaced by copy-on-write) part of the underlying shmem object out on swap.
 472"SwapPss" shows proportional swap share of this mapping. Unlike "Swap", this
 473does not take into account swapped out page of underlying shmem objects.
 474"Locked" indicates whether the mapping is locked in memory or not.
 475
 476"VmFlags" field deserves a separate description. This member represents the kernel
 477flags associated with the particular virtual memory area in two letter encoded
 478manner. The codes are the following:
 479    rd  - readable
 480    wr  - writeable
 481    ex  - executable
 482    sh  - shared
 483    mr  - may read
 484    mw  - may write
 485    me  - may execute
 486    ms  - may share
 487    gd  - stack segment growns down
 488    pf  - pure PFN range
 489    dw  - disabled write to the mapped file
 490    lo  - pages are locked in memory
 491    io  - memory mapped I/O area
 492    sr  - sequential read advise provided
 493    rr  - random read advise provided
 494    dc  - do not copy area on fork
 495    de  - do not expand area on remapping
 496    ac  - area is accountable
 497    nr  - swap space is not reserved for the area
 498    ht  - area uses huge tlb pages
 499    ar  - architecture specific flag
 500    dd  - do not include area into core dump
 501    sd  - soft-dirty flag
 502    mm  - mixed map area
 503    hg  - huge page advise flag
 504    nh  - no-huge page advise flag
 505    mg  - mergable advise flag
 506
 507Note that there is no guarantee that every flag and associated mnemonic will
 508be present in all further kernel releases. Things get changed, the flags may
 509be vanished or the reverse -- new added.
 510
 511This file is only present if the CONFIG_MMU kernel configuration option is
 512enabled.
 513
 514The /proc/PID/clear_refs is used to reset the PG_Referenced and ACCESSED/YOUNG
 515bits on both physical and virtual pages associated with a process, and the
 516soft-dirty bit on pte (see Documentation/vm/soft-dirty.txt for details).
 517To clear the bits for all the pages associated with the process
 518    > echo 1 > /proc/PID/clear_refs
 519
 520To clear the bits for the anonymous pages associated with the process
 521    > echo 2 > /proc/PID/clear_refs
 522
 523To clear the bits for the file mapped pages associated with the process
 524    > echo 3 > /proc/PID/clear_refs
 525
 526To clear the soft-dirty bit
 527    > echo 4 > /proc/PID/clear_refs
 528
 529To reset the peak resident set size ("high water mark") to the process's
 530current value:
 531    > echo 5 > /proc/PID/clear_refs
 532
 533Any other value written to /proc/PID/clear_refs will have no effect.
 534
 535The /proc/pid/pagemap gives the PFN, which can be used to find the pageflags
 536using /proc/kpageflags and number of times a page is mapped using
 537/proc/kpagecount. For detailed explanation, see Documentation/vm/pagemap.txt.
 538
 539The /proc/pid/numa_maps is an extension based on maps, showing the memory
 540locality and binding policy, as well as the memory usage (in pages) of
 541each mapping. The output follows a general format where mapping details get
 542summarized separated by blank spaces, one mapping per each file line:
 543
 544address   policy    mapping details
 545
 54600400000 default file=/usr/local/bin/app mapped=1 active=0 N3=1 kernelpagesize_kB=4
 54700600000 default file=/usr/local/bin/app anon=1 dirty=1 N3=1 kernelpagesize_kB=4
 5483206000000 default file=/lib64/ld-2.12.so mapped=26 mapmax=6 N0=24 N3=2 kernelpagesize_kB=4
 549320621f000 default file=/lib64/ld-2.12.so anon=1 dirty=1 N3=1 kernelpagesize_kB=4
 5503206220000 default file=/lib64/ld-2.12.so anon=1 dirty=1 N3=1 kernelpagesize_kB=4
 5513206221000 default anon=1 dirty=1 N3=1 kernelpagesize_kB=4
 5523206800000 default file=/lib64/libc-2.12.so mapped=59 mapmax=21 active=55 N0=41 N3=18 kernelpagesize_kB=4
 553320698b000 default file=/lib64/libc-2.12.so
 5543206b8a000 default file=/lib64/libc-2.12.so anon=2 dirty=2 N3=2 kernelpagesize_kB=4
 5553206b8e000 default file=/lib64/libc-2.12.so anon=1 dirty=1 N3=1 kernelpagesize_kB=4
 5563206b8f000 default anon=3 dirty=3 active=1 N3=3 kernelpagesize_kB=4
 5577f4dc10a2000 default anon=3 dirty=3 N3=3 kernelpagesize_kB=4
 5587f4dc10b4000 default anon=2 dirty=2 active=1 N3=2 kernelpagesize_kB=4
 5597f4dc1200000 default file=/anon_hugepage\040(deleted) huge anon=1 dirty=1 N3=1 kernelpagesize_kB=2048
 5607fff335f0000 default stack anon=3 dirty=3 N3=3 kernelpagesize_kB=4
 5617fff3369d000 default mapped=1 mapmax=35 active=0 N3=1 kernelpagesize_kB=4
 562
 563Where:
 564"address" is the starting address for the mapping;
 565"policy" reports the NUMA memory policy set for the mapping (see vm/numa_memory_policy.txt);
 566"mapping details" summarizes mapping data such as mapping type, page usage counters,
 567node locality page counters (N0 == node0, N1 == node1, ...) and the kernel page
 568size, in KB, that is backing the mapping up.
 569
 5701.2 Kernel data
 571---------------
 572
 573Similar to  the  process entries, the kernel data files give information about
 574the running kernel. The files used to obtain this information are contained in
 575/proc and  are  listed  in Table 1-5. Not all of these will be present in your
 576system. It  depends  on the kernel configuration and the loaded modules, which
 577files are there, and which are missing.
 578
 579Table 1-5: Kernel info in /proc
 580..............................................................................
 581 File        Content                                           
 582 apm         Advanced power management info                    
 583 buddyinfo   Kernel memory allocator information (see text)     (2.5)
 584 bus         Directory containing bus specific information     
 585 cmdline     Kernel command line                               
 586 cpuinfo     Info about the CPU                                
 587 devices     Available devices (block and character)           
 588 dma         Used DMS channels                                 
 589 filesystems Supported filesystems                             
 590 driver      Various drivers grouped here, currently rtc (2.4)
 591 execdomains Execdomains, related to security                   (2.4)
 592 fb          Frame Buffer devices                               (2.4)
 593 fs          File system parameters, currently nfs/exports      (2.4)
 594 ide         Directory containing info about the IDE subsystem 
 595 interrupts  Interrupt usage                                   
 596 iomem       Memory map                                         (2.4)
 597 ioports     I/O port usage                                    
 598 irq         Masks for irq to cpu affinity                      (2.4)(smp?)
 599 isapnp      ISA PnP (Plug&Play) Info                           (2.4)
 600 kcore       Kernel core image (can be ELF or A.OUT(deprecated in 2.4))   
 601 kmsg        Kernel messages                                   
 602 ksyms       Kernel symbol table                               
 603 loadavg     Load average of last 1, 5 & 15 minutes                
 604 locks       Kernel locks                                      
 605 meminfo     Memory info                                       
 606 misc        Miscellaneous                                     
 607 modules     List of loaded modules                            
 608 mounts      Mounted filesystems                               
 609 net         Networking info (see text)                        
 610 pagetypeinfo Additional page allocator information (see text)  (2.5)
 611 partitions  Table of partitions known to the system           
 612 pci         Deprecated info of PCI bus (new way -> /proc/bus/pci/,
 613             decoupled by lspci                                 (2.4)
 614 rtc         Real time clock                                   
 615 scsi        SCSI info (see text)                              
 616 slabinfo    Slab pool info                                    
 617 softirqs    softirq usage
 618 stat        Overall statistics                                
 619 swaps       Swap space utilization                            
 620 sys         See chapter 2                                     
 621 sysvipc     Info of SysVIPC Resources (msg, sem, shm)          (2.4)
 622 tty         Info of tty drivers
 623 uptime      Wall clock since boot, combined idle time of all cpus
 624 version     Kernel version                                    
 625 video       bttv info of video resources                       (2.4)
 626 vmallocinfo Show vmalloced areas
 627..............................................................................
 628
 629You can,  for  example,  check  which interrupts are currently in use and what
 630they are used for by looking in the file /proc/interrupts:
 631
 632  > cat /proc/interrupts 
 633             CPU0        
 634    0:    8728810          XT-PIC  timer 
 635    1:        895          XT-PIC  keyboard 
 636    2:          0          XT-PIC  cascade 
 637    3:     531695          XT-PIC  aha152x 
 638    4:    2014133          XT-PIC  serial 
 639    5:      44401          XT-PIC  pcnet_cs 
 640    8:          2          XT-PIC  rtc 
 641   11:          8          XT-PIC  i82365 
 642   12:     182918          XT-PIC  PS/2 Mouse 
 643   13:          1          XT-PIC  fpu 
 644   14:    1232265          XT-PIC  ide0 
 645   15:          7          XT-PIC  ide1 
 646  NMI:          0 
 647
 648In 2.4.* a couple of lines where added to this file LOC & ERR (this time is the
 649output of a SMP machine):
 650
 651  > cat /proc/interrupts 
 652
 653             CPU0       CPU1       
 654    0:    1243498    1214548    IO-APIC-edge  timer
 655    1:       8949       8958    IO-APIC-edge  keyboard
 656    2:          0          0          XT-PIC  cascade
 657    5:      11286      10161    IO-APIC-edge  soundblaster
 658    8:          1          0    IO-APIC-edge  rtc
 659    9:      27422      27407    IO-APIC-edge  3c503
 660   12:     113645     113873    IO-APIC-edge  PS/2 Mouse
 661   13:          0          0          XT-PIC  fpu
 662   14:      22491      24012    IO-APIC-edge  ide0
 663   15:       2183       2415    IO-APIC-edge  ide1
 664   17:      30564      30414   IO-APIC-level  eth0
 665   18:        177        164   IO-APIC-level  bttv
 666  NMI:    2457961    2457959 
 667  LOC:    2457882    2457881 
 668  ERR:       2155
 669
 670NMI is incremented in this case because every timer interrupt generates a NMI
 671(Non Maskable Interrupt) which is used by the NMI Watchdog to detect lockups.
 672
 673LOC is the local interrupt counter of the internal APIC of every CPU.
 674
 675ERR is incremented in the case of errors in the IO-APIC bus (the bus that
 676connects the CPUs in a SMP system. This means that an error has been detected,
 677the IO-APIC automatically retry the transmission, so it should not be a big
 678problem, but you should read the SMP-FAQ.
 679
 680In 2.6.2* /proc/interrupts was expanded again.  This time the goal was for
 681/proc/interrupts to display every IRQ vector in use by the system, not
 682just those considered 'most important'.  The new vectors are:
 683
 684  THR -- interrupt raised when a machine check threshold counter
 685  (typically counting ECC corrected errors of memory or cache) exceeds
 686  a configurable threshold.  Only available on some systems.
 687
 688  TRM -- a thermal event interrupt occurs when a temperature threshold
 689  has been exceeded for the CPU.  This interrupt may also be generated
 690  when the temperature drops back to normal.
 691
 692  SPU -- a spurious interrupt is some interrupt that was raised then lowered
 693  by some IO device before it could be fully processed by the APIC.  Hence
 694  the APIC sees the interrupt but does not know what device it came from.
 695  For this case the APIC will generate the interrupt with a IRQ vector
 696  of 0xff. This might also be generated by chipset bugs.
 697
 698  RES, CAL, TLB -- rescheduling, call and TLB flush interrupts are
 699  sent from one CPU to another per the needs of the OS.  Typically,
 700  their statistics are used by kernel developers and interested users to
 701  determine the occurrence of interrupts of the given type.
 702
 703The above IRQ vectors are displayed only when relevant.  For example,
 704the threshold vector does not exist on x86_64 platforms.  Others are
 705suppressed when the system is a uniprocessor.  As of this writing, only
 706i386 and x86_64 platforms support the new IRQ vector displays.
 707
 708Of some interest is the introduction of the /proc/irq directory to 2.4.
 709It could be used to set IRQ to CPU affinity, this means that you can "hook" an
 710IRQ to only one CPU, or to exclude a CPU of handling IRQs. The contents of the
 711irq subdir is one subdir for each IRQ, and two files; default_smp_affinity and
 712prof_cpu_mask.
 713
 714For example 
 715  > ls /proc/irq/
 716  0  10  12  14  16  18  2  4  6  8  prof_cpu_mask
 717  1  11  13  15  17  19  3  5  7  9  default_smp_affinity
 718  > ls /proc/irq/0/
 719  smp_affinity
 720
 721smp_affinity is a bitmask, in which you can specify which CPUs can handle the
 722IRQ, you can set it by doing:
 723
 724  > echo 1 > /proc/irq/10/smp_affinity
 725
 726This means that only the first CPU will handle the IRQ, but you can also echo
 7275 which means that only the first and fourth CPU can handle the IRQ.
 728
 729The contents of each smp_affinity file is the same by default:
 730
 731  > cat /proc/irq/0/smp_affinity
 732  ffffffff
 733
 734There is an alternate interface, smp_affinity_list which allows specifying
 735a cpu range instead of a bitmask:
 736
 737  > cat /proc/irq/0/smp_affinity_list
 738  1024-1031
 739
 740The default_smp_affinity mask applies to all non-active IRQs, which are the
 741IRQs which have not yet been allocated/activated, and hence which lack a
 742/proc/irq/[0-9]* directory.
 743
 744The node file on an SMP system shows the node to which the device using the IRQ
 745reports itself as being attached. This hardware locality information does not
 746include information about any possible driver locality preference.
 747
 748prof_cpu_mask specifies which CPUs are to be profiled by the system wide
 749profiler. Default value is ffffffff (all cpus if there are only 32 of them).
 750
 751The way IRQs are routed is handled by the IO-APIC, and it's Round Robin
 752between all the CPUs which are allowed to handle it. As usual the kernel has
 753more info than you and does a better job than you, so the defaults are the
 754best choice for almost everyone.  [Note this applies only to those IO-APIC's
 755that support "Round Robin" interrupt distribution.]
 756
 757There are  three  more  important subdirectories in /proc: net, scsi, and sys.
 758The general  rule  is  that  the  contents,  or  even  the  existence of these
 759directories, depend  on your kernel configuration. If SCSI is not enabled, the
 760directory scsi  may  not  exist. The same is true with the net, which is there
 761only when networking support is present in the running kernel.
 762
 763The slabinfo  file  gives  information  about  memory usage at the slab level.
 764Linux uses  slab  pools for memory management above page level in version 2.2.
 765Commonly used  objects  have  their  own  slab  pool (such as network buffers,
 766directory cache, and so on).
 767
 768..............................................................................
 769
 770> cat /proc/buddyinfo
 771
 772Node 0, zone      DMA      0      4      5      4      4      3 ...
 773Node 0, zone   Normal      1      0      0      1    101      8 ...
 774Node 0, zone  HighMem      2      0      0      1      1      0 ...
 775
 776External fragmentation is a problem under some workloads, and buddyinfo is a
 777useful tool for helping diagnose these problems.  Buddyinfo will give you a 
 778clue as to how big an area you can safely allocate, or why a previous
 779allocation failed.
 780
 781Each column represents the number of pages of a certain order which are 
 782available.  In this case, there are 0 chunks of 2^0*PAGE_SIZE available in 
 783ZONE_DMA, 4 chunks of 2^1*PAGE_SIZE in ZONE_DMA, 101 chunks of 2^4*PAGE_SIZE 
 784available in ZONE_NORMAL, etc... 
 785
 786More information relevant to external fragmentation can be found in
 787pagetypeinfo.
 788
 789> cat /proc/pagetypeinfo
 790Page block order: 9
 791Pages per block:  512
 792
 793Free pages count per migrate type at order       0      1      2      3      4      5      6      7      8      9     10
 794Node    0, zone      DMA, type    Unmovable      0      0      0      1      1      1      1      1      1      1      0
 795Node    0, zone      DMA, type  Reclaimable      0      0      0      0      0      0      0      0      0      0      0
 796Node    0, zone      DMA, type      Movable      1      1      2      1      2      1      1      0      1      0      2
 797Node    0, zone      DMA, type      Reserve      0      0      0      0      0      0      0      0      0      1      0
 798Node    0, zone      DMA, type      Isolate      0      0      0      0      0      0      0      0      0      0      0
 799Node    0, zone    DMA32, type    Unmovable    103     54     77      1      1      1     11      8      7      1      9
 800Node    0, zone    DMA32, type  Reclaimable      0      0      2      1      0      0      0      0      1      0      0
 801Node    0, zone    DMA32, type      Movable    169    152    113     91     77     54     39     13      6      1    452
 802Node    0, zone    DMA32, type      Reserve      1      2      2      2      2      0      1      1      1      1      0
 803Node    0, zone    DMA32, type      Isolate      0      0      0      0      0      0      0      0      0      0      0
 804
 805Number of blocks type     Unmovable  Reclaimable      Movable      Reserve      Isolate
 806Node 0, zone      DMA            2            0            5            1            0
 807Node 0, zone    DMA32           41            6          967            2            0
 808
 809Fragmentation avoidance in the kernel works by grouping pages of different
 810migrate types into the same contiguous regions of memory called page blocks.
 811A page block is typically the size of the default hugepage size e.g. 2MB on
 812X86-64. By keeping pages grouped based on their ability to move, the kernel
 813can reclaim pages within a page block to satisfy a high-order allocation.
 814
 815The pagetypinfo begins with information on the size of a page block. It
 816then gives the same type of information as buddyinfo except broken down
 817by migrate-type and finishes with details on how many page blocks of each
 818type exist.
 819
 820If min_free_kbytes has been tuned correctly (recommendations made by hugeadm
 821from libhugetlbfs https://github.com/libhugetlbfs/libhugetlbfs/), one can
 822make an estimate of the likely number of huge pages that can be allocated
 823at a given point in time. All the "Movable" blocks should be allocatable
 824unless memory has been mlock()'d. Some of the Reclaimable blocks should
 825also be allocatable although a lot of filesystem metadata may have to be
 826reclaimed to achieve this.
 827
 828..............................................................................
 829
 830meminfo:
 831
 832Provides information about distribution and utilization of memory.  This
 833varies by architecture and compile options.  The following is from a
 83416GB PIII, which has highmem enabled.  You may not have all of these fields.
 835
 836> cat /proc/meminfo
 837
 838MemTotal:     16344972 kB
 839MemFree:      13634064 kB
 840MemAvailable: 14836172 kB
 841Buffers:          3656 kB
 842Cached:        1195708 kB
 843SwapCached:          0 kB
 844Active:         891636 kB
 845Inactive:      1077224 kB
 846HighTotal:    15597528 kB
 847HighFree:     13629632 kB
 848LowTotal:       747444 kB
 849LowFree:          4432 kB
 850SwapTotal:           0 kB
 851SwapFree:            0 kB
 852Dirty:             968 kB
 853Writeback:           0 kB
 854AnonPages:      861800 kB
 855Mapped:         280372 kB
 856Shmem:             644 kB
 857Slab:           284364 kB
 858SReclaimable:   159856 kB
 859SUnreclaim:     124508 kB
 860PageTables:      24448 kB
 861NFS_Unstable:        0 kB
 862Bounce:              0 kB
 863WritebackTmp:        0 kB
 864CommitLimit:   7669796 kB
 865Committed_AS:   100056 kB
 866VmallocTotal:   112216 kB
 867VmallocUsed:       428 kB
 868VmallocChunk:   111088 kB
 869AnonHugePages:   49152 kB
 870
 871    MemTotal: Total usable ram (i.e. physical ram minus a few reserved
 872              bits and the kernel binary code)
 873     MemFree: The sum of LowFree+HighFree
 874MemAvailable: An estimate of how much memory is available for starting new
 875              applications, without swapping. Calculated from MemFree,
 876              SReclaimable, the size of the file LRU lists, and the low
 877              watermarks in each zone.
 878              The estimate takes into account that the system needs some
 879              page cache to function well, and that not all reclaimable
 880              slab will be reclaimable, due to items being in use. The
 881              impact of those factors will vary from system to system.
 882     Buffers: Relatively temporary storage for raw disk blocks
 883              shouldn't get tremendously large (20MB or so)
 884      Cached: in-memory cache for files read from the disk (the
 885              pagecache).  Doesn't include SwapCached
 886  SwapCached: Memory that once was swapped out, is swapped back in but
 887              still also is in the swapfile (if memory is needed it
 888              doesn't need to be swapped out AGAIN because it is already
 889              in the swapfile. This saves I/O)
 890      Active: Memory that has been used more recently and usually not
 891              reclaimed unless absolutely necessary.
 892    Inactive: Memory which has been less recently used.  It is more
 893              eligible to be reclaimed for other purposes
 894   HighTotal:
 895    HighFree: Highmem is all memory above ~860MB of physical memory
 896              Highmem areas are for use by userspace programs, or
 897              for the pagecache.  The kernel must use tricks to access
 898              this memory, making it slower to access than lowmem.
 899    LowTotal:
 900     LowFree: Lowmem is memory which can be used for everything that
 901              highmem can be used for, but it is also available for the
 902              kernel's use for its own data structures.  Among many
 903              other things, it is where everything from the Slab is
 904              allocated.  Bad things happen when you're out of lowmem.
 905   SwapTotal: total amount of swap space available
 906    SwapFree: Memory which has been evicted from RAM, and is temporarily
 907              on the disk
 908       Dirty: Memory which is waiting to get written back to the disk
 909   Writeback: Memory which is actively being written back to the disk
 910   AnonPages: Non-file backed pages mapped into userspace page tables
 911AnonHugePages: Non-file backed huge pages mapped into userspace page tables
 912      Mapped: files which have been mmaped, such as libraries
 913       Shmem: Total memory used by shared memory (shmem) and tmpfs
 914        Slab: in-kernel data structures cache
 915SReclaimable: Part of Slab, that might be reclaimed, such as caches
 916  SUnreclaim: Part of Slab, that cannot be reclaimed on memory pressure
 917  PageTables: amount of memory dedicated to the lowest level of page
 918              tables.
 919NFS_Unstable: NFS pages sent to the server, but not yet committed to stable
 920              storage
 921      Bounce: Memory used for block device "bounce buffers"
 922WritebackTmp: Memory used by FUSE for temporary writeback buffers
 923 CommitLimit: Based on the overcommit ratio ('vm.overcommit_ratio'),
 924              this is the total amount of  memory currently available to
 925              be allocated on the system. This limit is only adhered to
 926              if strict overcommit accounting is enabled (mode 2 in
 927              'vm.overcommit_memory').
 928              The CommitLimit is calculated with the following formula:
 929              CommitLimit = ([total RAM pages] - [total huge TLB pages]) *
 930                             overcommit_ratio / 100 + [total swap pages]
 931              For example, on a system with 1G of physical RAM and 7G
 932              of swap with a `vm.overcommit_ratio` of 30 it would
 933              yield a CommitLimit of 7.3G.
 934              For more details, see the memory overcommit documentation
 935              in vm/overcommit-accounting.
 936Committed_AS: The amount of memory presently allocated on the system.
 937              The committed memory is a sum of all of the memory which
 938              has been allocated by processes, even if it has not been
 939              "used" by them as of yet. A process which malloc()'s 1G
 940              of memory, but only touches 300M of it will show up as
 941              using 1G. This 1G is memory which has been "committed" to
 942              by the VM and can be used at any time by the allocating
 943              application. With strict overcommit enabled on the system
 944              (mode 2 in 'vm.overcommit_memory'),allocations which would
 945              exceed the CommitLimit (detailed above) will not be permitted.
 946              This is useful if one needs to guarantee that processes will
 947              not fail due to lack of memory once that memory has been
 948              successfully allocated.
 949VmallocTotal: total size of vmalloc memory area
 950 VmallocUsed: amount of vmalloc area which is used
 951VmallocChunk: largest contiguous block of vmalloc area which is free
 952
 953..............................................................................
 954
 955vmallocinfo:
 956
 957Provides information about vmalloced/vmaped areas. One line per area,
 958containing the virtual address range of the area, size in bytes,
 959caller information of the creator, and optional information depending
 960on the kind of area :
 961
 962 pages=nr    number of pages
 963 phys=addr   if a physical address was specified
 964 ioremap     I/O mapping (ioremap() and friends)
 965 vmalloc     vmalloc() area
 966 vmap        vmap()ed pages
 967 user        VM_USERMAP area
 968 vpages      buffer for pages pointers was vmalloced (huge area)
 969 N<node>=nr  (Only on NUMA kernels)
 970             Number of pages allocated on memory node <node>
 971
 972> cat /proc/vmallocinfo
 9730xffffc20000000000-0xffffc20000201000 2101248 alloc_large_system_hash+0x204 ...
 974  /0x2c0 pages=512 vmalloc N0=128 N1=128 N2=128 N3=128
 9750xffffc20000201000-0xffffc20000302000 1052672 alloc_large_system_hash+0x204 ...
 976  /0x2c0 pages=256 vmalloc N0=64 N1=64 N2=64 N3=64
 9770xffffc20000302000-0xffffc20000304000    8192 acpi_tb_verify_table+0x21/0x4f...
 978  phys=7fee8000 ioremap
 9790xffffc20000304000-0xffffc20000307000   12288 acpi_tb_verify_table+0x21/0x4f...
 980  phys=7fee7000 ioremap
 9810xffffc2000031d000-0xffffc2000031f000    8192 init_vdso_vars+0x112/0x210
 9820xffffc2000031f000-0xffffc2000032b000   49152 cramfs_uncompress_init+0x2e ...
 983  /0x80 pages=11 vmalloc N0=3 N1=3 N2=2 N3=3
 9840xffffc2000033a000-0xffffc2000033d000   12288 sys_swapon+0x640/0xac0      ...
 985  pages=2 vmalloc N1=2
 9860xffffc20000347000-0xffffc2000034c000   20480 xt_alloc_table_info+0xfe ...
 987  /0x130 [x_tables] pages=4 vmalloc N0=4
 9880xffffffffa0000000-0xffffffffa000f000   61440 sys_init_module+0xc27/0x1d00 ...
 989   pages=14 vmalloc N2=14
 9900xffffffffa000f000-0xffffffffa0014000   20480 sys_init_module+0xc27/0x1d00 ...
 991   pages=4 vmalloc N1=4
 9920xffffffffa0014000-0xffffffffa0017000   12288 sys_init_module+0xc27/0x1d00 ...
 993   pages=2 vmalloc N1=2
 9940xffffffffa0017000-0xffffffffa0022000   45056 sys_init_module+0xc27/0x1d00 ...
 995   pages=10 vmalloc N0=10
 996
 997..............................................................................
 998
 999softirqs:
1000
1001Provides counts of softirq handlers serviced since boot time, for each cpu.
1002
1003> cat /proc/softirqs
1004                CPU0       CPU1       CPU2       CPU3
1005      HI:          0          0          0          0
1006   TIMER:      27166      27120      27097      27034
1007  NET_TX:          0          0          0         17
1008  NET_RX:         42          0          0         39
1009   BLOCK:          0          0        107       1121
1010 TASKLET:          0          0          0        290
1011   SCHED:      27035      26983      26971      26746
1012 HRTIMER:          0          0          0          0
1013     RCU:       1678       1769       2178       2250
1014
1015
10161.3 IDE devices in /proc/ide
1017----------------------------
1018
1019The subdirectory /proc/ide contains information about all IDE devices of which
1020the kernel  is  aware.  There is one subdirectory for each IDE controller, the
1021file drivers  and a link for each IDE device, pointing to the device directory
1022in the controller specific subtree.
1023
1024The file  drivers  contains general information about the drivers used for the
1025IDE devices:
1026
1027  > cat /proc/ide/drivers
1028  ide-cdrom version 4.53
1029  ide-disk version 1.08
1030
1031More detailed  information  can  be  found  in  the  controller  specific
1032subdirectories. These  are  named  ide0,  ide1  and  so  on.  Each  of  these
1033directories contains the files shown in table 1-6.
1034
1035
1036Table 1-6: IDE controller info in  /proc/ide/ide?
1037..............................................................................
1038 File    Content                                 
1039 channel IDE channel (0 or 1)                    
1040 config  Configuration (only for PCI/IDE bridge) 
1041 mate    Mate name                               
1042 model   Type/Chipset of IDE controller          
1043..............................................................................
1044
1045Each device  connected  to  a  controller  has  a separate subdirectory in the
1046controllers directory.  The  files  listed in table 1-7 are contained in these
1047directories.
1048
1049
1050Table 1-7: IDE device information
1051..............................................................................
1052 File             Content                                    
1053 cache            The cache                                  
1054 capacity         Capacity of the medium (in 512Byte blocks) 
1055 driver           driver and version                         
1056 geometry         physical and logical geometry              
1057 identify         device identify block                      
1058 media            media type                                 
1059 model            device identifier                          
1060 settings         device setup                               
1061 smart_thresholds IDE disk management thresholds             
1062 smart_values     IDE disk management values                 
1063..............................................................................
1064
1065The most  interesting  file is settings. This file contains a nice overview of
1066the drive parameters:
1067
1068  # cat /proc/ide/ide0/hda/settings 
1069  name                    value           min             max             mode 
1070  ----                    -----           ---             ---             ---- 
1071  bios_cyl                526             0               65535           rw 
1072  bios_head               255             0               255             rw 
1073  bios_sect               63              0               63              rw 
1074  breada_readahead        4               0               127             rw 
1075  bswap                   0               0               1               r 
1076  file_readahead          72              0               2097151         rw 
1077  io_32bit                0               0               3               rw 
1078  keepsettings            0               0               1               rw 
1079  max_kb_per_request      122             1               127             rw 
1080  multcount               0               0               8               rw 
1081  nice1                   1               0               1               rw 
1082  nowerr                  0               0               1               rw 
1083  pio_mode                write-only      0               255             w 
1084  slow                    0               0               1               rw 
1085  unmaskirq               0               0               1               rw 
1086  using_dma               0               0               1               rw 
1087
1088
10891.4 Networking info in /proc/net
1090--------------------------------
1091
1092The subdirectory  /proc/net  follows  the  usual  pattern. Table 1-8 shows the
1093additional values  you  get  for  IP  version 6 if you configure the kernel to
1094support this. Table 1-9 lists the files and their meaning.
1095
1096
1097Table 1-8: IPv6 info in /proc/net
1098..............................................................................
1099 File       Content                                               
1100 udp6       UDP sockets (IPv6)                                    
1101 tcp6       TCP sockets (IPv6)                                    
1102 raw6       Raw device statistics (IPv6)                          
1103 igmp6      IP multicast addresses, which this host joined (IPv6) 
1104 if_inet6   List of IPv6 interface addresses                      
1105 ipv6_route Kernel routing table for IPv6                         
1106 rt6_stats  Global IPv6 routing tables statistics                 
1107 sockstat6  Socket statistics (IPv6)                              
1108 snmp6      Snmp data (IPv6)                                      
1109..............................................................................
1110
1111
1112Table 1-9: Network info in /proc/net
1113..............................................................................
1114 File          Content                                                         
1115 arp           Kernel  ARP table                                               
1116 dev           network devices with statistics                                 
1117 dev_mcast     the Layer2 multicast groups a device is listening too
1118               (interface index, label, number of references, number of bound
1119               addresses). 
1120 dev_stat      network device status                                           
1121 ip_fwchains   Firewall chain linkage                                          
1122 ip_fwnames    Firewall chain names                                            
1123 ip_masq       Directory containing the masquerading tables                    
1124 ip_masquerade Major masquerading table                                        
1125 netstat       Network statistics                                              
1126 raw           raw device statistics                                           
1127 route         Kernel routing table                                            
1128 rpc           Directory containing rpc info                                   
1129 rt_cache      Routing cache                                                   
1130 snmp          SNMP data                                                       
1131 sockstat      Socket statistics                                               
1132 tcp           TCP  sockets                                                    
1133 udp           UDP sockets                                                     
1134 unix          UNIX domain sockets                                             
1135 wireless      Wireless interface data (Wavelan etc)                           
1136 igmp          IP multicast addresses, which this host joined                  
1137 psched        Global packet scheduler parameters.                             
1138 netlink       List of PF_NETLINK sockets                                      
1139 ip_mr_vifs    List of multicast virtual interfaces                            
1140 ip_mr_cache   List of multicast routing cache                                 
1141..............................................................................
1142
1143You can  use  this  information  to see which network devices are available in
1144your system and how much traffic was routed over those devices:
1145
1146  > cat /proc/net/dev 
1147  Inter-|Receive                                                   |[... 
1148   face |bytes    packets errs drop fifo frame compressed multicast|[... 
1149      lo:  908188   5596     0    0    0     0          0         0 [...         
1150    ppp0:15475140  20721   410    0    0   410          0         0 [...  
1151    eth0:  614530   7085     0    0    0     0          0         1 [... 
1152   
1153  ...] Transmit 
1154  ...] bytes    packets errs drop fifo colls carrier compressed 
1155  ...]  908188     5596    0    0    0     0       0          0 
1156  ...] 1375103    17405    0    0    0     0       0          0 
1157  ...] 1703981     5535    0    0    0     3       0          0 
1158
1159In addition, each Channel Bond interface has its own directory.  For
1160example, the bond0 device will have a directory called /proc/net/bond0/.
1161It will contain information that is specific to that bond, such as the
1162current slaves of the bond, the link status of the slaves, and how
1163many times the slaves link has failed.
1164
11651.5 SCSI info
1166-------------
1167
1168If you  have  a  SCSI  host adapter in your system, you'll find a subdirectory
1169named after  the driver for this adapter in /proc/scsi. You'll also see a list
1170of all recognized SCSI devices in /proc/scsi:
1171
1172  >cat /proc/scsi/scsi 
1173  Attached devices: 
1174  Host: scsi0 Channel: 00 Id: 00 Lun: 00 
1175    Vendor: IBM      Model: DGHS09U          Rev: 03E0 
1176    Type:   Direct-Access                    ANSI SCSI revision: 03 
1177  Host: scsi0 Channel: 00 Id: 06 Lun: 00 
1178    Vendor: PIONEER  Model: CD-ROM DR-U06S   Rev: 1.04 
1179    Type:   CD-ROM                           ANSI SCSI revision: 02 
1180
1181
1182The directory  named  after  the driver has one file for each adapter found in
1183the system.  These  files  contain information about the controller, including
1184the used  IRQ  and  the  IO  address range. The amount of information shown is
1185dependent on  the adapter you use. The example shows the output for an Adaptec
1186AHA-2940 SCSI adapter:
1187
1188  > cat /proc/scsi/aic7xxx/0 
1189   
1190  Adaptec AIC7xxx driver version: 5.1.19/3.2.4 
1191  Compile Options: 
1192    TCQ Enabled By Default : Disabled 
1193    AIC7XXX_PROC_STATS     : Disabled 
1194    AIC7XXX_RESET_DELAY    : 5 
1195  Adapter Configuration: 
1196             SCSI Adapter: Adaptec AHA-294X Ultra SCSI host adapter 
1197                             Ultra Wide Controller 
1198      PCI MMAPed I/O Base: 0xeb001000 
1199   Adapter SEEPROM Config: SEEPROM found and used. 
1200        Adaptec SCSI BIOS: Enabled 
1201                      IRQ: 10 
1202                     SCBs: Active 0, Max Active 2, 
1203                           Allocated 15, HW 16, Page 255 
1204               Interrupts: 160328 
1205        BIOS Control Word: 0x18b6 
1206     Adapter Control Word: 0x005b 
1207     Extended Translation: Enabled 
1208  Disconnect Enable Flags: 0xffff 
1209       Ultra Enable Flags: 0x0001 
1210   Tag Queue Enable Flags: 0x0000 
1211  Ordered Queue Tag Flags: 0x0000 
1212  Default Tag Queue Depth: 8 
1213      Tagged Queue By Device array for aic7xxx host instance 0: 
1214        {255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255} 
1215      Actual queue depth per device for aic7xxx host instance 0: 
1216        {1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1} 
1217  Statistics: 
1218  (scsi0:0:0:0) 
1219    Device using Wide/Sync transfers at 40.0 MByte/sec, offset 8 
1220    Transinfo settings: current(12/8/1/0), goal(12/8/1/0), user(12/15/1/0) 
1221    Total transfers 160151 (74577 reads and 85574 writes) 
1222  (scsi0:0:6:0) 
1223    Device using Narrow/Sync transfers at 5.0 MByte/sec, offset 15 
1224    Transinfo settings: current(50/15/0/0), goal(50/15/0/0), user(50/15/0/0) 
1225    Total transfers 0 (0 reads and 0 writes) 
1226
1227
12281.6 Parallel port info in /proc/parport
1229---------------------------------------
1230
1231The directory  /proc/parport  contains information about the parallel ports of
1232your system.  It  has  one  subdirectory  for  each port, named after the port
1233number (0,1,2,...).
1234
1235These directories contain the four files shown in Table 1-10.
1236
1237
1238Table 1-10: Files in /proc/parport
1239..............................................................................
1240 File      Content                                                             
1241 autoprobe Any IEEE-1284 device ID information that has been acquired.         
1242 devices   list of the device drivers using that port. A + will appear by the
1243           name of the device currently using the port (it might not appear
1244           against any). 
1245 hardware  Parallel port's base address, IRQ line and DMA channel.             
1246 irq       IRQ that parport is using for that port. This is in a separate
1247           file to allow you to alter it by writing a new value in (IRQ
1248           number or none). 
1249..............................................................................
1250
12511.7 TTY info in /proc/tty
1252-------------------------
1253
1254Information about  the  available  and actually used tty's can be found in the
1255directory /proc/tty.You'll  find  entries  for drivers and line disciplines in
1256this directory, as shown in Table 1-11.
1257
1258
1259Table 1-11: Files in /proc/tty
1260..............................................................................
1261 File          Content                                        
1262 drivers       list of drivers and their usage                
1263 ldiscs        registered line disciplines                    
1264 driver/serial usage statistic and status of single tty lines 
1265..............................................................................
1266
1267To see  which  tty's  are  currently in use, you can simply look into the file
1268/proc/tty/drivers:
1269
1270  > cat /proc/tty/drivers 
1271  pty_slave            /dev/pts      136   0-255 pty:slave 
1272  pty_master           /dev/ptm      128   0-255 pty:master 
1273  pty_slave            /dev/ttyp       3   0-255 pty:slave 
1274  pty_master           /dev/pty        2   0-255 pty:master 
1275  serial               /dev/cua        5   64-67 serial:callout 
1276  serial               /dev/ttyS       4   64-67 serial 
1277  /dev/tty0            /dev/tty0       4       0 system:vtmaster 
1278  /dev/ptmx            /dev/ptmx       5       2 system 
1279  /dev/console         /dev/console    5       1 system:console 
1280  /dev/tty             /dev/tty        5       0 system:/dev/tty 
1281  unknown              /dev/tty        4    1-63 console 
1282
1283
12841.8 Miscellaneous kernel statistics in /proc/stat
1285-------------------------------------------------
1286
1287Various pieces   of  information about  kernel activity  are  available in the
1288/proc/stat file.  All  of  the numbers reported  in  this file are  aggregates
1289since the system first booted.  For a quick look, simply cat the file:
1290
1291  > cat /proc/stat
1292  cpu  2255 34 2290 22625563 6290 127 456 0 0 0
1293  cpu0 1132 34 1441 11311718 3675 127 438 0 0 0
1294  cpu1 1123 0 849 11313845 2614 0 18 0 0 0
1295  intr 114930548 113199788 3 0 5 263 0 4 [... lots more numbers ...]
1296  ctxt 1990473
1297  btime 1062191376
1298  processes 2915
1299  procs_running 1
1300  procs_blocked 0
1301  softirq 183433 0 21755 12 39 1137 231 21459 2263
1302
1303The very first  "cpu" line aggregates the  numbers in all  of the other "cpuN"
1304lines.  These numbers identify the amount of time the CPU has spent performing
1305different kinds of work.  Time units are in USER_HZ (typically hundredths of a
1306second).  The meanings of the columns are as follows, from left to right:
1307
1308- user: normal processes executing in user mode
1309- nice: niced processes executing in user mode
1310- system: processes executing in kernel mode
1311- idle: twiddling thumbs
1312- iowait: waiting for I/O to complete
1313- irq: servicing interrupts
1314- softirq: servicing softirqs
1315- steal: involuntary wait
1316- guest: running a normal guest
1317- guest_nice: running a niced guest
1318
1319The "intr" line gives counts of interrupts  serviced since boot time, for each
1320of the  possible system interrupts.   The first  column  is the  total of  all
1321interrupts serviced  including  unnumbered  architecture specific  interrupts;
1322each  subsequent column is the  total for that particular numbered interrupt.
1323Unnumbered interrupts are not shown, only summed into the total.
1324
1325The "ctxt" line gives the total number of context switches across all CPUs.
1326
1327The "btime" line gives  the time at which the  system booted, in seconds since
1328the Unix epoch.
1329
1330The "processes" line gives the number  of processes and threads created, which
1331includes (but  is not limited  to) those  created by  calls to the  fork() and
1332clone() system calls.
1333
1334The "procs_running" line gives the total number of threads that are
1335running or ready to run (i.e., the total number of runnable threads).
1336
1337The   "procs_blocked" line gives  the  number of  processes currently blocked,
1338waiting for I/O to complete.
1339
1340The "softirq" line gives counts of softirqs serviced since boot time, for each
1341of the possible system softirqs. The first column is the total of all
1342softirqs serviced; each subsequent column is the total for that particular
1343softirq.
1344
1345
13461.9 Ext4 file system parameters
1347-------------------------------
1348
1349Information about mounted ext4 file systems can be found in
1350/proc/fs/ext4.  Each mounted filesystem will have a directory in
1351/proc/fs/ext4 based on its device name (i.e., /proc/fs/ext4/hdc or
1352/proc/fs/ext4/dm-0).   The files in each per-device directory are shown
1353in Table 1-12, below.
1354
1355Table 1-12: Files in /proc/fs/ext4/<devname>
1356..............................................................................
1357 File            Content                                        
1358 mb_groups       details of multiblock allocator buddy cache of free blocks
1359..............................................................................
1360
13612.0 /proc/consoles
1362------------------
1363Shows registered system console lines.
1364
1365To see which character device lines are currently used for the system console
1366/dev/console, you may simply look into the file /proc/consoles:
1367
1368  > cat /proc/consoles
1369  tty0                 -WU (ECp)       4:7
1370  ttyS0                -W- (Ep)        4:64
1371
1372The columns are:
1373
1374  device               name of the device
1375  operations           R = can do read operations
1376                       W = can do write operations
1377                       U = can do unblank
1378  flags                E = it is enabled
1379                       C = it is preferred console
1380                       B = it is primary boot console
1381                       p = it is used for printk buffer
1382                       b = it is not a TTY but a Braille device
1383                       a = it is safe to use when cpu is offline
1384  major:minor          major and minor number of the device separated by a colon
1385
1386------------------------------------------------------------------------------
1387Summary
1388------------------------------------------------------------------------------
1389The /proc file system serves information about the running system. It not only
1390allows access to process data but also allows you to request the kernel status
1391by reading files in the hierarchy.
1392
1393The directory  structure  of /proc reflects the types of information and makes
1394it easy, if not obvious, where to look for specific data.
1395------------------------------------------------------------------------------
1396
1397------------------------------------------------------------------------------
1398CHAPTER 2: MODIFYING SYSTEM PARAMETERS
1399------------------------------------------------------------------------------
1400
1401------------------------------------------------------------------------------
1402In This Chapter
1403------------------------------------------------------------------------------
1404* Modifying kernel parameters by writing into files found in /proc/sys
1405* Exploring the files which modify certain parameters
1406* Review of the /proc/sys file tree
1407------------------------------------------------------------------------------
1408
1409
1410A very  interesting part of /proc is the directory /proc/sys. This is not only
1411a source  of  information,  it also allows you to change parameters within the
1412kernel. Be  very  careful  when attempting this. You can optimize your system,
1413but you  can  also  cause  it  to  crash.  Never  alter kernel parameters on a
1414production system.  Set  up  a  development machine and test to make sure that
1415everything works  the  way  you want it to. You may have no alternative but to
1416reboot the machine once an error has been made.
1417
1418To change  a  value,  simply  echo  the new value into the file. An example is
1419given below  in the section on the file system data. You need to be root to do
1420this. You  can  create  your  own  boot script to perform this every time your
1421system boots.
1422
1423The files  in /proc/sys can be used to fine tune and monitor miscellaneous and
1424general things  in  the operation of the Linux kernel. Since some of the files
1425can inadvertently  disrupt  your  system,  it  is  advisable  to  read  both
1426documentation and  source  before actually making adjustments. In any case, be
1427very careful  when  writing  to  any  of these files. The entries in /proc may
1428change slightly between the 2.1.* and the 2.2 kernel, so if there is any doubt
1429review the kernel documentation in the directory /usr/src/linux/Documentation.
1430This chapter  is  heavily  based  on the documentation included in the pre 2.2
1431kernels, and became part of it in version 2.2.1 of the Linux kernel.
1432
1433Please see: Documentation/sysctl/ directory for descriptions of these
1434entries.
1435
1436------------------------------------------------------------------------------
1437Summary
1438------------------------------------------------------------------------------
1439Certain aspects  of  kernel  behavior  can be modified at runtime, without the
1440need to  recompile  the kernel, or even to reboot the system. The files in the
1441/proc/sys tree  can  not only be read, but also modified. You can use the echo
1442command to write value into these files, thereby changing the default settings
1443of the kernel.
1444------------------------------------------------------------------------------
1445
1446------------------------------------------------------------------------------
1447CHAPTER 3: PER-PROCESS PARAMETERS
1448------------------------------------------------------------------------------
1449
14503.1 /proc/<pid>/oom_adj & /proc/<pid>/oom_score_adj- Adjust the oom-killer score
1451--------------------------------------------------------------------------------
1452
1453These file can be used to adjust the badness heuristic used to select which
1454process gets killed in out of memory conditions.
1455
1456The badness heuristic assigns a value to each candidate task ranging from 0
1457(never kill) to 1000 (always kill) to determine which process is targeted.  The
1458units are roughly a proportion along that range of allowed memory the process
1459may allocate from based on an estimation of its current memory and swap use.
1460For example, if a task is using all allowed memory, its badness score will be
14611000.  If it is using half of its allowed memory, its score will be 500.
1462
1463There is an additional factor included in the badness score: the current memory
1464and swap usage is discounted by 3% for root processes.
1465
1466The amount of "allowed" memory depends on the context in which the oom killer
1467was called.  If it is due to the memory assigned to the allocating task's cpuset
1468being exhausted, the allowed memory represents the set of mems assigned to that
1469cpuset.  If it is due to a mempolicy's node(s) being exhausted, the allowed
1470memory represents the set of mempolicy nodes.  If it is due to a memory
1471limit (or swap limit) being reached, the allowed memory is that configured
1472limit.  Finally, if it is due to the entire system being out of memory, the
1473allowed memory represents all allocatable resources.
1474
1475The value of /proc/<pid>/oom_score_adj is added to the badness score before it
1476is used to determine which task to kill.  Acceptable values range from -1000
1477(OOM_SCORE_ADJ_MIN) to +1000 (OOM_SCORE_ADJ_MAX).  This allows userspace to
1478polarize the preference for oom killing either by always preferring a certain
1479task or completely disabling it.  The lowest possible value, -1000, is
1480equivalent to disabling oom killing entirely for that task since it will always
1481report a badness score of 0.
1482
1483Consequently, it is very simple for userspace to define the amount of memory to
1484consider for each task.  Setting a /proc/<pid>/oom_score_adj value of +500, for
1485example, is roughly equivalent to allowing the remainder of tasks sharing the
1486same system, cpuset, mempolicy, or memory controller resources to use at least
148750% more memory.  A value of -500, on the other hand, would be roughly
1488equivalent to discounting 50% of the task's allowed memory from being considered
1489as scoring against the task.
1490
1491For backwards compatibility with previous kernels, /proc/<pid>/oom_adj may also
1492be used to tune the badness score.  Its acceptable values range from -16
1493(OOM_ADJUST_MIN) to +15 (OOM_ADJUST_MAX) and a special value of -17
1494(OOM_DISABLE) to disable oom killing entirely for that task.  Its value is
1495scaled linearly with /proc/<pid>/oom_score_adj.
1496
1497The value of /proc/<pid>/oom_score_adj may be reduced no lower than the last
1498value set by a CAP_SYS_RESOURCE process. To reduce the value any lower
1499requires CAP_SYS_RESOURCE.
1500
1501Caveat: when a parent task is selected, the oom killer will sacrifice any first
1502generation children with separate address spaces instead, if possible.  This
1503avoids servers and important system daemons from being killed and loses the
1504minimal amount of work.
1505
1506
15073.2 /proc/<pid>/oom_score - Display current oom-killer score
1508-------------------------------------------------------------
1509
1510This file can be used to check the current score used by the oom-killer is for
1511any given <pid>. Use it together with /proc/<pid>/oom_score_adj to tune which
1512process should be killed in an out-of-memory situation.
1513
1514
15153.3  /proc/<pid>/io - Display the IO accounting fields
1516-------------------------------------------------------
1517
1518This file contains IO statistics for each running process
1519
1520Example
1521-------
1522
1523test:/tmp # dd if=/dev/zero of=/tmp/test.dat &
1524[1] 3828
1525
1526test:/tmp # cat /proc/3828/io
1527rchar: 323934931
1528wchar: 323929600
1529syscr: 632687
1530syscw: 632675
1531read_bytes: 0
1532write_bytes: 323932160
1533cancelled_write_bytes: 0
1534
1535
1536Description
1537-----------
1538
1539rchar
1540-----
1541
1542I/O counter: chars read
1543The number of bytes which this task has caused to be read from storage. This
1544is simply the sum of bytes which this process passed to read() and pread().
1545It includes things like tty IO and it is unaffected by whether or not actual
1546physical disk IO was required (the read might have been satisfied from
1547pagecache)
1548
1549
1550wchar
1551-----
1552
1553I/O counter: chars written
1554The number of bytes which this task has caused, or shall cause to be written
1555to disk. Similar caveats apply here as with rchar.
1556
1557
1558syscr
1559-----
1560
1561I/O counter: read syscalls
1562Attempt to count the number of read I/O operations, i.e. syscalls like read()
1563and pread().
1564
1565
1566syscw
1567-----
1568
1569I/O counter: write syscalls
1570Attempt to count the number of write I/O operations, i.e. syscalls like
1571write() and pwrite().
1572
1573
1574read_bytes
1575----------
1576
1577I/O counter: bytes read
1578Attempt to count the number of bytes which this process really did cause to
1579be fetched from the storage layer. Done at the submit_bio() level, so it is
1580accurate for block-backed filesystems. <please add status regarding NFS and
1581CIFS at a later time>
1582
1583
1584write_bytes
1585-----------
1586
1587I/O counter: bytes written
1588Attempt to count the number of bytes which this process caused to be sent to
1589the storage layer. This is done at page-dirtying time.
1590
1591
1592cancelled_write_bytes
1593---------------------
1594
1595The big inaccuracy here is truncate. If a process writes 1MB to a file and
1596then deletes the file, it will in fact perform no writeout. But it will have
1597been accounted as having caused 1MB of write.
1598In other words: The number of bytes which this process caused to not happen,
1599by truncating pagecache. A task can cause "negative" IO too. If this task
1600truncates some dirty pagecache, some IO which another task has been accounted
1601for (in its write_bytes) will not be happening. We _could_ just subtract that
1602from the truncating task's write_bytes, but there is information loss in doing
1603that.
1604
1605
1606Note
1607----
1608
1609At its current implementation state, this is a bit racy on 32-bit machines: if
1610process A reads process B's /proc/pid/io while process B is updating one of
1611those 64-bit counters, process A could see an intermediate result.
1612
1613
1614More information about this can be found within the taskstats documentation in
1615Documentation/accounting.
1616
16173.4 /proc/<pid>/coredump_filter - Core dump filtering settings
1618---------------------------------------------------------------
1619When a process is dumped, all anonymous memory is written to a core file as
1620long as the size of the core file isn't limited. But sometimes we don't want
1621to dump some memory segments, for example, huge shared memory or DAX.
1622Conversely, sometimes we want to save file-backed memory segments into a core
1623file, not only the individual files.
1624
1625/proc/<pid>/coredump_filter allows you to customize which memory segments
1626will be dumped when the <pid> process is dumped. coredump_filter is a bitmask
1627of memory types. If a bit of the bitmask is set, memory segments of the
1628corresponding memory type are dumped, otherwise they are not dumped.
1629
1630The following 9 memory types are supported:
1631  - (bit 0) anonymous private memory
1632  - (bit 1) anonymous shared memory
1633  - (bit 2) file-backed private memory
1634  - (bit 3) file-backed shared memory
1635  - (bit 4) ELF header pages in file-backed private memory areas (it is
1636            effective only if the bit 2 is cleared)
1637  - (bit 5) hugetlb private memory
1638  - (bit 6) hugetlb shared memory
1639  - (bit 7) DAX private memory
1640  - (bit 8) DAX shared memory
1641
1642  Note that MMIO pages such as frame buffer are never dumped and vDSO pages
1643  are always dumped regardless of the bitmask status.
1644
1645  Note that bits 0-4 don't affect hugetlb or DAX memory. hugetlb memory is
1646  only affected by bit 5-6, and DAX is only affected by bits 7-8.
1647
1648The default value of coredump_filter is 0x33; this means all anonymous memory
1649segments, ELF header pages and hugetlb private memory are dumped.
1650
1651If you don't want to dump all shared memory segments attached to pid 1234,
1652write 0x31 to the process's proc file.
1653
1654  $ echo 0x31 > /proc/1234/coredump_filter
1655
1656When a new process is created, the process inherits the bitmask status from its
1657parent. It is useful to set up coredump_filter before the program runs.
1658For example:
1659
1660  $ echo 0x7 > /proc/self/coredump_filter
1661  $ ./some_program
1662
16633.5     /proc/<pid>/mountinfo - Information about mounts
1664--------------------------------------------------------
1665
1666This file contains lines of the form:
1667
166836 35 98:0 /mnt1 /mnt2 rw,noatime master:1 - ext3 /dev/root rw,errors=continue
1669(1)(2)(3)   (4)   (5)      (6)      (7)   (8) (9)   (10)         (11)
1670
1671(1) mount ID:  unique identifier of the mount (may be reused after umount)
1672(2) parent ID:  ID of parent (or of self for the top of the mount tree)
1673(3) major:minor:  value of st_dev for files on filesystem
1674(4) root:  root of the mount within the filesystem
1675(5) mount point:  mount point relative to the process's root
1676(6) mount options:  per mount options
1677(7) optional fields:  zero or more fields of the form "tag[:value]"
1678(8) separator:  marks the end of the optional fields
1679(9) filesystem type:  name of filesystem of the form "type[.subtype]"
1680(10) mount source:  filesystem specific information or "none"
1681(11) super options:  per super block options
1682
1683Parsers should ignore all unrecognised optional fields.  Currently the
1684possible optional fields are:
1685
1686shared:X  mount is shared in peer group X
1687master:X  mount is slave to peer group X
1688propagate_from:X  mount is slave and receives propagation from peer group X (*)
1689unbindable  mount is unbindable
1690
1691(*) X is the closest dominant peer group under the process's root.  If
1692X is the immediate master of the mount, or if there's no dominant peer
1693group under the same root, then only the "master:X" field is present
1694and not the "propagate_from:X" field.
1695
1696For more information on mount propagation see:
1697
1698  Documentation/filesystems/sharedsubtree.txt
1699
1700
17013.6     /proc/<pid>/comm  & /proc/<pid>/task/<tid>/comm
1702--------------------------------------------------------
1703These files provide a method to access a tasks comm value. It also allows for
1704a task to set its own or one of its thread siblings comm value. The comm value
1705is limited in size compared to the cmdline value, so writing anything longer
1706then the kernel's TASK_COMM_LEN (currently 16 chars) will result in a truncated
1707comm value.
1708
1709
17103.7     /proc/<pid>/task/<tid>/children - Information about task children
1711-------------------------------------------------------------------------
1712This file provides a fast way to retrieve first level children pids
1713of a task pointed by <pid>/<tid> pair. The format is a space separated
1714stream of pids.
1715
1716Note the "first level" here -- if a child has own children they will
1717not be listed here, one needs to read /proc/<children-pid>/task/<tid>/children
1718to obtain the descendants.
1719
1720Since this interface is intended to be fast and cheap it doesn't
1721guarantee to provide precise results and some children might be
1722skipped, especially if they've exited right after we printed their
1723pids, so one need to either stop or freeze processes being inspected
1724if precise results are needed.
1725
1726
17273.8     /proc/<pid>/fdinfo/<fd> - Information about opened file
1728---------------------------------------------------------------
1729This file provides information associated with an opened file. The regular
1730files have at least three fields -- 'pos', 'flags' and mnt_id. The 'pos'
1731represents the current offset of the opened file in decimal form [see lseek(2)
1732for details], 'flags' denotes the octal O_xxx mask the file has been
1733created with [see open(2) for details] and 'mnt_id' represents mount ID of
1734the file system containing the opened file [see 3.5 /proc/<pid>/mountinfo
1735for details].
1736
1737A typical output is
1738
1739        pos:    0
1740        flags:  0100002
1741        mnt_id: 19
1742
1743All locks associated with a file descriptor are shown in its fdinfo too.
1744
1745lock:       1: FLOCK  ADVISORY  WRITE 359 00:13:11691 0 EOF
1746
1747The files such as eventfd, fsnotify, signalfd, epoll among the regular pos/flags
1748pair provide additional information particular to the objects they represent.
1749
1750        Eventfd files
1751        ~~~~~~~~~~~~~
1752        pos:    0
1753        flags:  04002
1754        mnt_id: 9
1755        eventfd-count:  5a
1756
1757        where 'eventfd-count' is hex value of a counter.
1758
1759        Signalfd files
1760        ~~~~~~~~~~~~~~
1761        pos:    0
1762        flags:  04002
1763        mnt_id: 9
1764        sigmask:        0000000000000200
1765
1766        where 'sigmask' is hex value of the signal mask associated
1767        with a file.
1768
1769        Epoll files
1770        ~~~~~~~~~~~
1771        pos:    0
1772        flags:  02
1773        mnt_id: 9
1774        tfd:        5 events:       1d data: ffffffffffffffff
1775
1776        where 'tfd' is a target file descriptor number in decimal form,
1777        'events' is events mask being watched and the 'data' is data
1778        associated with a target [see epoll(7) for more details].
1779
1780        Fsnotify files
1781        ~~~~~~~~~~~~~~
1782        For inotify files the format is the following
1783
1784        pos:    0
1785        flags:  02000000
1786        inotify wd:3 ino:9e7e sdev:800013 mask:800afce ignored_mask:0 fhandle-bytes:8 fhandle-type:1 f_handle:7e9e0000640d1b6d
1787
1788        where 'wd' is a watch descriptor in decimal form, ie a target file
1789        descriptor number, 'ino' and 'sdev' are inode and device where the
1790        target file resides and the 'mask' is the mask of events, all in hex
1791        form [see inotify(7) for more details].
1792
1793        If the kernel was built with exportfs support, the path to the target
1794        file is encoded as a file handle.  The file handle is provided by three
1795        fields 'fhandle-bytes', 'fhandle-type' and 'f_handle', all in hex
1796        format.
1797
1798        If the kernel is built without exportfs support the file handle won't be
1799        printed out.
1800
1801        If there is no inotify mark attached yet the 'inotify' line will be omitted.
1802
1803        For fanotify files the format is
1804
1805        pos:    0
1806        flags:  02
1807        mnt_id: 9
1808        fanotify flags:10 event-flags:0
1809        fanotify mnt_id:12 mflags:40 mask:38 ignored_mask:40000003
1810        fanotify ino:4f969 sdev:800013 mflags:0 mask:3b ignored_mask:40000000 fhandle-bytes:8 fhandle-type:1 f_handle:69f90400c275b5b4
1811
1812        where fanotify 'flags' and 'event-flags' are values used in fanotify_init
1813        call, 'mnt_id' is the mount point identifier, 'mflags' is the value of
1814        flags associated with mark which are tracked separately from events
1815        mask. 'ino', 'sdev' are target inode and device, 'mask' is the events
1816        mask and 'ignored_mask' is the mask of events which are to be ignored.
1817        All in hex format. Incorporation of 'mflags', 'mask' and 'ignored_mask'
1818        does provide information about flags and mask used in fanotify_mark
1819        call [see fsnotify manpage for details].
1820
1821        While the first three lines are mandatory and always printed, the rest is
1822        optional and may be omitted if no marks created yet.
1823
1824        Timerfd files
1825        ~~~~~~~~~~~~~
1826
1827        pos:    0
1828        flags:  02
1829        mnt_id: 9
1830        clockid: 0
1831        ticks: 0
1832        settime flags: 01
1833        it_value: (0, 49406829)
1834        it_interval: (1, 0)
1835
1836        where 'clockid' is the clock type and 'ticks' is the number of the timer expirations
1837        that have occurred [see timerfd_create(2) for details]. 'settime flags' are
1838        flags in octal form been used to setup the timer [see timerfd_settime(2) for
1839        details]. 'it_value' is remaining time until the timer exiration.
1840        'it_interval' is the interval for the timer. Note the timer might be set up
1841        with TIMER_ABSTIME option which will be shown in 'settime flags', but 'it_value'
1842        still exhibits timer's remaining time.
1843
18443.9     /proc/<pid>/map_files - Information about memory mapped files
1845---------------------------------------------------------------------
1846This directory contains symbolic links which represent memory mapped files
1847the process is maintaining.  Example output:
1848
1849     | lr-------- 1 root root 64 Jan 27 11:24 333c600000-333c620000 -> /usr/lib64/ld-2.18.so
1850     | lr-------- 1 root root 64 Jan 27 11:24 333c81f000-333c820000 -> /usr/lib64/ld-2.18.so
1851     | lr-------- 1 root root 64 Jan 27 11:24 333c820000-333c821000 -> /usr/lib64/ld-2.18.so
1852     | ...
1853     | lr-------- 1 root root 64 Jan 27 11:24 35d0421000-35d0422000 -> /usr/lib64/libselinux.so.1
1854     | lr-------- 1 root root 64 Jan 27 11:24 400000-41a000 -> /usr/bin/ls
1855
1856The name of a link represents the virtual memory bounds of a mapping, i.e.
1857vm_area_struct::vm_start-vm_area_struct::vm_end.
1858
1859The main purpose of the map_files is to retrieve a set of memory mapped
1860files in a fast way instead of parsing /proc/<pid>/maps or
1861/proc/<pid>/smaps, both of which contain many more records.  At the same
1862time one can open(2) mappings from the listings of two processes and
1863comparing their inode numbers to figure out which anonymous memory areas
1864are actually shared.
1865
18663.10    /proc/<pid>/timerslack_ns - Task timerslack value
1867---------------------------------------------------------
1868This file provides the value of the task's timerslack value in nanoseconds.
1869This value specifies a amount of time that normal timers may be deferred
1870in order to coalesce timers and avoid unnecessary wakeups.
1871
1872This allows a task's interactivity vs power consumption trade off to be
1873adjusted.
1874
1875Writing 0 to the file will set the tasks timerslack to the default value.
1876
1877Valid values are from 0 - ULLONG_MAX
1878
1879An application setting the value must have PTRACE_MODE_ATTACH_FSCREDS level
1880permissions on the task specified to change its timerslack_ns value.
1881
1882
1883------------------------------------------------------------------------------
1884Configuring procfs
1885------------------------------------------------------------------------------
1886
18874.1     Mount options
1888---------------------
1889
1890The following mount options are supported:
1891
1892        hidepid=        Set /proc/<pid>/ access mode.
1893        gid=            Set the group authorized to learn processes information.
1894
1895hidepid=0 means classic mode - everybody may access all /proc/<pid>/ directories
1896(default).
1897
1898hidepid=1 means users may not access any /proc/<pid>/ directories but their
1899own.  Sensitive files like cmdline, sched*, status are now protected against
1900other users.  This makes it impossible to learn whether any user runs
1901specific program (given the program doesn't reveal itself by its behaviour).
1902As an additional bonus, as /proc/<pid>/cmdline is unaccessible for other users,
1903poorly written programs passing sensitive information via program arguments are
1904now protected against local eavesdroppers.
1905
1906hidepid=2 means hidepid=1 plus all /proc/<pid>/ will be fully invisible to other
1907users.  It doesn't mean that it hides a fact whether a process with a specific
1908pid value exists (it can be learned by other means, e.g. by "kill -0 $PID"),
1909but it hides process' uid and gid, which may be learned by stat()'ing
1910/proc/<pid>/ otherwise.  It greatly complicates an intruder's task of gathering
1911information about running processes, whether some daemon runs with elevated
1912privileges, whether other user runs some sensitive program, whether other users
1913run any program at all, etc.
1914
1915gid= defines a group authorized to learn processes information otherwise
1916prohibited by hidepid=.  If you use some daemon like identd which needs to learn
1917information about processes information, just add identd to this group.
1918
lxr.linux.no kindly hosted by Redpill Linpro AS, provider of Linux consulting and operations services since 1995.